|
|
Linha 1: |
Linha 1: |
| + | <ads media=goo5/> |
| === RHCS(Red Hat Cluster Suite) quorum disk === | | === RHCS(Red Hat Cluster Suite) quorum disk === |
| <div style="margin-top: 0px; margin-right: 0px; margin-bottom: 0.75em; margin-left: 0px; line-height: 1.6em; color: rgb(51, 51, 51); font-family: Verdana, serif; font-size: 13px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: left; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255); " id="post-body-1979843398346145118" class="post-body entry-content">The last post "[http://honglus.blogspot.com/2011/05/rhcsred-hat-cluster-suite-io-fencing.html RHCS I/O fencing]" is about dealing with split-brain situation, in which cluster members lost heartbeat communication and each believe it is legitimate to write data to the shared storage.<span class="Apple-converted-space"> </span><br>'''Methods to deal with split-brain situation:<span class="Apple-converted-space"> </span><br>'''1. Redundant heartbeat path<span class="Apple-converted-space"> </span><br>network port communication plus serial port communication<span class="Apple-converted-space"> </span><br>2. I/O fencing<span class="Apple-converted-space"> </span><br>Remaining nodes separate failed node from its storage either by shutdown/reboot power port or storage port<span class="Apple-converted-space"> </span><br>3. Quorum disk<span class="Apple-converted-space"> </span><br>Quorum disk is a kind of I/O fencing, but the reboot action is executed by failed node's own quorum daemon. It also has<span class="Apple-converted-space"> </span>'''additional feature''': contributing vote to cluster. if you want the last standing node to keep the multiple-nodes cluster running, quorum disk appears to be the only solution.<br>'''RHCS (Red Hat Cluster Suite) Quorum disk facts<span class="Apple-converted-space"> </span><br>'''- A shared block device (SCSI/iSCSI/FC..), Device size requirement is approximately 10MiB<span class="Apple-converted-space"> </span><br>- Supports maximum 16 nodes, nodes id must be sequentially ordered<span class="Apple-converted-space"> </span><br>- Quorum disk can contribute votes. In multiple nodes cluster, together with quorum vote, the last standing node can still keep the cluster running<span class="Apple-converted-space"> </span><br>- single node votes+1 <=Quorum's disk vote < nodes total votes<span class="Apple-converted-space"> </span><br>- The failure of the shared quorum disk won’t result in cluster failure, as long as Quorum's disk vote < nodes total votes<span class="Apple-converted-space"> </span><br>- each node write its own health information in its own region, the health is determined by external checking program such as "ping"<br>'''Setup Quorum disk'''<br><pre>#initialise quorum disk once in any node | | <div style="margin-top: 0px; margin-right: 0px; margin-bottom: 0.75em; margin-left: 0px; line-height: 1.6em; color: rgb(51, 51, 51); font-family: Verdana, serif; font-size: 13px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: left; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255); " id="post-body-1979843398346145118" class="post-body entry-content">The last post "[http://honglus.blogspot.com/2011/05/rhcsred-hat-cluster-suite-io-fencing.html RHCS I/O fencing]" is about dealing with split-brain situation, in which cluster members lost heartbeat communication and each believe it is legitimate to write data to the shared storage.<span class="Apple-converted-space"> </span><br>'''Methods to deal with split-brain situation:<span class="Apple-converted-space"> </span><br>'''1. Redundant heartbeat path<span class="Apple-converted-space"> </span><br>network port communication plus serial port communication<span class="Apple-converted-space"> </span><br>2. I/O fencing<span class="Apple-converted-space"> </span><br>Remaining nodes separate failed node from its storage either by shutdown/reboot power port or storage port<span class="Apple-converted-space"> </span><br>3. Quorum disk<span class="Apple-converted-space"> </span><br>Quorum disk is a kind of I/O fencing, but the reboot action is executed by failed node's own quorum daemon. It also has<span class="Apple-converted-space"> </span>'''additional feature''': contributing vote to cluster. if you want the last standing node to keep the multiple-nodes cluster running, quorum disk appears to be the only solution.<br>'''RHCS (Red Hat Cluster Suite) Quorum disk facts<span class="Apple-converted-space"> </span><br>'''- A shared block device (SCSI/iSCSI/FC..), Device size requirement is approximately 10MiB<span class="Apple-converted-space"> </span><br>- Supports maximum 16 nodes, nodes id must be sequentially ordered<span class="Apple-converted-space"> </span><br>- Quorum disk can contribute votes. In multiple nodes cluster, together with quorum vote, the last standing node can still keep the cluster running<span class="Apple-converted-space"> </span><br>- single node votes+1 <=Quorum's disk vote < nodes total votes<span class="Apple-converted-space"> </span><br>- The failure of the shared quorum disk won’t result in cluster failure, as long as Quorum's disk vote < nodes total votes<span class="Apple-converted-space"> </span><br>- each node write its own health information in its own region, the health is determined by external checking program such as "ping"<br>'''Setup Quorum disk'''<br><pre>#initialise quorum disk once in any node |
Linha 52: |
Linha 53: |
| </pre><pre style="background-color: rgb(251, 251, 251); font-family: consolas, 'Courier New', courier, monospace; font-size: 12px; margin-top: 0em; margin-right: 0em; margin-bottom: 0em; margin-left: 0em; width: 650px; ">Quorum: 5 | | </pre><pre style="background-color: rgb(251, 251, 251); font-family: consolas, 'Courier New', courier, monospace; font-size: 12px; margin-top: 0em; margin-right: 0em; margin-bottom: 0em; margin-left: 0em; width: 650px; ">Quorum: 5 |
| </pre><pre style="background-color: rgb(251, 251, 251); font-family: consolas, 'Courier New', courier, monospace; font-size: 12px; margin-top: 0em; margin-right: 0em; margin-bottom: 0em; margin-left: 0em; width: 650px; ">..</pre></pre></div><div style="margin-top: 0.75em; margin-right: 0px; margin-bottom: 0.75em; margin-left: 0px; color: rgb(153, 153, 153); text-transform: uppercase; letter-spacing: 0.1em; font: normal normal normal 78%/normal 'Trebuchet MS', Trebuchet, Arial, Verdana, sans-serif; line-height: 1.4em; font-style: normal; font-variant: normal; font-weight: normal; orphans: 2; text-align: left; text-indent: 0px; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255); " class="post-footer"><div class="post-footer-line post-footer-line-1"><span class="post-author vcard">POSTED BY<span class="Apple-converted-space"> </span><span class="fn">HONGLUS</span><span class="Apple-converted-space"> </span></span><span class="post-timestamp">AT<span class="Apple-converted-space"> </span>[http://honglus.blogspot.com/2011/05/rhcsred-hat-cluster-suite-quorum-disk.html <abbr style="border-top-style: none; border-right-style: none; border-bottom-style: none; border-left-style: none; border-width: initial; border-color: initial; " title="2011-05-07T17:12:00+10:00" class="published">5:12 PM</abbr>]<span class="Apple-converted-space"> </span></span><span class="reaction-buttons" /><span class="star-ratings" /><span class="post-comment-link" /><span class="post-backlinks post-comment-link" /><span class="post-icons"><span class="item-action">[http://www.blogger.com/email-post.g?blogID=1822605347830719087&postID=1979843398346145118 [[Image:]]<span class="Apple-converted-space"> </span>]</span></span></div></div> | | </pre><pre style="background-color: rgb(251, 251, 251); font-family: consolas, 'Courier New', courier, monospace; font-size: 12px; margin-top: 0em; margin-right: 0em; margin-bottom: 0em; margin-left: 0em; width: 650px; ">..</pre></pre></div><div style="margin-top: 0.75em; margin-right: 0px; margin-bottom: 0.75em; margin-left: 0px; color: rgb(153, 153, 153); text-transform: uppercase; letter-spacing: 0.1em; font: normal normal normal 78%/normal 'Trebuchet MS', Trebuchet, Arial, Verdana, sans-serif; line-height: 1.4em; font-style: normal; font-variant: normal; font-weight: normal; orphans: 2; text-align: left; text-indent: 0px; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255); " class="post-footer"><div class="post-footer-line post-footer-line-1"><span class="post-author vcard">POSTED BY<span class="Apple-converted-space"> </span><span class="fn">HONGLUS</span><span class="Apple-converted-space"> </span></span><span class="post-timestamp">AT<span class="Apple-converted-space"> </span>[http://honglus.blogspot.com/2011/05/rhcsred-hat-cluster-suite-quorum-disk.html <abbr style="border-top-style: none; border-right-style: none; border-bottom-style: none; border-left-style: none; border-width: initial; border-color: initial; " title="2011-05-07T17:12:00+10:00" class="published">5:12 PM</abbr>]<span class="Apple-converted-space"> </span></span><span class="reaction-buttons" /><span class="star-ratings" /><span class="post-comment-link" /><span class="post-backlinks post-comment-link" /><span class="post-icons"><span class="item-action">[http://www.blogger.com/email-post.g?blogID=1822605347830719087&postID=1979843398346145118 [[Image:]]<span class="Apple-converted-space"> </span>]</span></span></div></div> |
| + | |
| + | <ads media=goo3/> |
The last post "
RHCS I/O fencing" is about dealing with split-brain situation, in which cluster members lost heartbeat communication and each believe it is legitimate to write data to the shared storage.
Methods to deal with split-brain situation:
1. Redundant heartbeat path
network port communication plus serial port communication
2. I/O fencing
Remaining nodes separate failed node from its storage either by shutdown/reboot power port or storage port
3. Quorum disk
Quorum disk is a kind of I/O fencing, but the reboot action is executed by failed node's own quorum daemon. It also has
additional feature: contributing vote to cluster. if you want the last standing node to keep the multiple-nodes cluster running, quorum disk appears to be the only solution.
RHCS (Red Hat Cluster Suite) Quorum disk facts
- A shared block device (SCSI/iSCSI/FC..), Device size requirement is approximately 10MiB
- Supports maximum 16 nodes, nodes id must be sequentially ordered
- Quorum disk can contribute votes. In multiple nodes cluster, together with quorum vote, the last standing node can still keep the cluster running
- single node votes+1 <=Quorum's disk vote < nodes total votes
- The failure of the shared quorum disk won’t result in cluster failure, as long as Quorum's disk vote < nodes total votes
- each node write its own health information in its own region, the health is determined by external checking program such as "ping"
Setup Quorum disk#initialise quorum disk once in any node
mkqdisk -c /dev/sdx -l myqdisk
Add quorum disk to cluster Use luci or system-config-cluster to add quorum disk, following is the result xml file
<pre style="background-color: rgb(251, 251, 251); font-family: consolas, 'Courier New', courier, monospace; font-size: 12px; margin-top: 0em; margin-right: 0em; margin-bottom: 0em; margin-left: 0em; width: 650px; "><clusternodes>
<clusternode name="<span style="color: rgb(139, 0, 0); ">station1.example.com</span>" nodeid="<span style="color: rgb(139, 0, 0); ">1</span>" votes="<span style="color: rgb(139, 0, 0); ">2</span>">
<fence/>
</clusternode>
<clusternode name="<span style="color: rgb(139, 0, 0); ">station2.example.com</span>" nodeid="<span style="color: rgb(139, 0, 0); ">2</span>" votes="<span style="color: rgb(139, 0, 0); ">2</span>">
<fence/>
</clusternode>
<clusternode name="<span style="color: rgb(139, 0, 0); ">station3.example.com</span>" nodeid="<span style="color: rgb(139, 0, 0); ">3</span>" votes="<span style="color: rgb(139, 0, 0); ">2</span>">
<fence/>
</clusternode>
</clusternodes>
<span style="color: green; ">#expected votes =9=(nodes total votes + quorum disk votes) = (2+2+2+3) </span>
<cman expected_votes="<span style="color: rgb(139, 0, 0); ">9</span>"/>
<span style="color: green; ">#Health check result is writen to quorum disk every 2 secs</span>
<span style="color: green; ">#if health check fails over 5 tko, 10 (2*5) secs, the node is rebooted by quorum daemon</span>
<span style="color: green; ">#Each heuristic check is run very 2 secs and earn 1 score,if shell script return is 0</span>
<quorumd interval="<span style="color: rgb(139, 0, 0); ">2</span>" label="<span style="color: rgb(139, 0, 0); ">myqdisk</span>" min_score="<span style="color: rgb(139, 0, 0); ">2</span>" tko="<span style="color: rgb(139, 0, 0); ">5</span>" votes="<span style="color: rgb(139, 0, 0); ">3</span>">
<heuristic interval="<span style="color: rgb(139, 0, 0); ">2</span>" program="<span style="color: rgb(139, 0, 0); ">ping -c1 -t1 192.168.1.60</span>" score="<span style="color: rgb(139, 0, 0); ">1</span>"/>
<heuristic interval="<span style="color: rgb(139, 0, 0); ">2</span>" program="<span style="color: rgb(139, 0, 0); ">ping -c1 -t1 192.168.1.254</span>" score="<span style="color: rgb(139, 0, 0); ">1</span>"/>
</quorumd>
</pre>
Start quorum disk daemonThe daemon is also one of daemons automatically started by cman
service qdiskd startCheck quorum disk information<pre style="background-color: rgb(251, 251, 251); font-family: consolas, 'Courier New', courier, monospace; font-size: 12px; margin-top: 0em; margin-right: 0em; margin-bottom: 0em; margin-left: 0em; width: 650px; ">$ mkqdisk -L -d
mkqdisk v0.6.0
/dev/disk/by-id/scsi-1IET_00010002:
/dev/disk/by-uuid/55fbf858-df75-493b-a764-5640be5a9b46:
/dev/sdc:
Magic: eb7a62c2
Label: myqdisk
Created: Sat May 7 05:56:35 2011
Host: station2.example.com
Kernel Sector Size: 512
Recorded Sector Size: 512
Status block <span style="color: blue; ">for</span> node 1
Last updated by node 1
Last updated on Sat May 7 15:09:37 2011
State: Master
Flags: 0000
Score: 0/0
Average Cycle speed: 0.001500 seconds
Last Cycle speed: 0.000000 seconds
Incarnation: 4dc4d1764dc4d176
Status block <span style="color: blue; ">for</span> node 2
Last updated by node 2
Last updated on Sun May 8 01:09:38 2011
State: Running
Flags: 0000
Score: 0/0
Average Cycle speed: 0.001000 seconds
Last Cycle speed: 0.000000 seconds
Incarnation: 4dc55e164dc55e16
Status block <span style="color: blue; ">for</span> node 3
Last updated by node 3
Last updated on Sat May 7 15:09:38 2011
State: Running
Flags: 0000
Score: 0/0
Average Cycle speed: 0.001500 seconds
Last Cycle speed: 0.000000 seconds
Incarnation: 4dc4d2f04dc4d2f0
</pre>
The cluster is still running with last node standingPlease note Total votes=quorum votes=5=2+3, if quorum disk vote is less than (node votes+1), the cluster wouldn’t have survived
<pre style="background-color: rgb(251, 251, 251); font-family: consolas, 'Courier New', courier, monospace; font-size: 12px; margin-top: 0em; margin-right: 0em; margin-bottom: 0em; margin-left: 0em; width: 650px; ">$cman_tool status
..
Nodes: 1
Expected votes: 9
Quorum device votes: 3
Total votes: 5
Quorum: 5
..
</pre>